Description

This article explains the following syslog messages:

  • 'NF Chip RODR packet drop'
  • 'RODR offset overflow count incremented'

Symptoms

Syslog Messages:

Jan 24 23:38:34 : %PFE-3: fpc5 Cmerror Op Set: NF Chip: NFCHIP(0) RODR packet drop
Jan 24 23:38:34  : %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop
Jan 24 23:38:34  : %PFE-3: fpc5 NFAB(5/1): RODR offset overflow count incremented (1)
Jun 24 23:38:34  alarmd[1644]: %DAEMON-5: Alarm set: FPC color=RED, class=CHASSIS, reason=FPC 5 Major Errors

Jun 24 23:38:34  craftd[1645]: %DAEMON-5:  Major alarm set, FPC 5 Major Errors

Solution

Check the FPC Type:

user@host > show chassis hardware 
<snip>
FPC 5            REV 07   710-013560   xxxxx            E2-FPC Type 3

 

Check Alarm:

user@host> show chassis alarms no-forwarding

1 alarms currently active
Alarm time               Class  Description
2020-01-24 23:38:34 UTC  Major  FPC 5 Major Errors

 

Check if FPC is online:

user@host> show chassis fpc detail
Slot 5 information:
  State                               Online
  Temperature                      53 degrees C / 127 degrees F
  Total CPU DRAM                 1024 MB
  Total SRAM                       72 MB
  Total SDRAM                    1280 MB


If FPC is offline, you can directly go to restarting the FPC to bring it back online. 

Detailed logs:

Jan 20 09:12:01  %PFE-3: fpc5 NF Chip: NFCHIP(1) RODR packet drop <-- Make note of NFCHIP number. Here, it is NFCHIP 1
Jan 20 09:13:16  %PFE-3: fpc5 NFAB(5/1): RODR offset overflow count incremented (1)
Jan 20 09:21:20  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop - clear 
Jan 20 09:34:14  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop 
Jan 20 09:36:20  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop - clear 
Jan 20 09:49:14  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop 
Jan 20 09:51:20  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop - clear 
Jan 20 10:04:17  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop 
Jan 20 10:06:23  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop - clear 
Jan 20 10:19:13  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop 
Jan 20 10:21:20  %PFE-3: fpc5 NF Chip:NFCHIP(1) RODR packet drop – clear

Typically, when an FPC receives packets in the form of cells, it reassembles it. So at the time of reassembly when the FPC is congested and when it is unable to receive the cells further, it will drop the packet without reordering (RODR) it, then we see the logs as above.

Check RODR statistics on NCHIP 1 for FPC 5:

user@host> request pfe execute command "show nchip 1 rodr" target fpc5

NFAB(1/0) Nin RODR statistics: (110713827)

Statistics            Total   Rate/sec  Peak Rate    Err Sec
cell_timeout            307          0         64          5
ra_overflow            1293          0         23        627
summary_pkt_drop       1324          0          6       1079
err_cell                  0          0          0          0
offset_overflow           7          0          1          7

The above command helps us understand the rate and error rate at which the occurrence happened. 

In most cases, alarm clears on its own and the errors stop appearing. If the error does not clear or FPC is offline, please restart the FPC.

FPC Restart:

user@host>  request chassis fpc restart slot  slot-number

 

Post restart, if the error still remains or FPC still remains offline, contact JTAC for further investigation.
Collect the following logs for JTAC analysis:

show pfe statistics traffic fpc x  <--- Replace x with affected slot
show chassis sibs
show chassis fabric fpcs
show chassis fabric sibs
show chassis fabric topology
request pfe execute command  "show hsl2 statistics crc " target fpcX  <-- Replace x with affacted FPC. Run 3-4 times at 5 mins interval

Modification History

07/12/2024 added a recent case from a T1600 router running JUNOS 14.2R4-S9